SoylentNews

Freeman writes:

https://arstechnica.com/science/2023/11/ai-outperforms-conventional-weather-forecasting-for-the-first-time-google-study/

On Tuesday, the peer-reviewed journal Science published a study that shows how an AI meteorology model from Google DeepMind called GraphCast has significantly outperformed conventional weather forecasting methods in predicting global weather conditions up to 10 days in advance. The achievement suggests that future weather forecasting may become far more accurate, reports The Washington Post and Financial Times.

In the study, GraphCast demonstrated superior performance over the world's leading conventional system, operated by the European Centre for Medium-range Weather Forecasts (ECMWF). In a comprehensive evaluation, GraphCast outperformed ECMWF's system in 90 percent of 1,380 metrics, including temperature, pressure, wind speed and direction, and humidity at various atmospheric levels.

Original Submission

upstart writes on behalf of aristarchus:

AMD Readies 8-Core Ryzen 7 5700X3D and 6-Core Ryzen 5 5500X3D With 96MB L3 Cache:

As per information from chi11eddog, a reliable source for hardware leaks regarding CPU and motherboard launches, AMD is reportedly in the process of developing at least two additional CPUs in the Ryzen 5000X3D series. One of these is a more affordable 8-core model known as the Ryzen 7 5700X3D, anticipated to boast a base clock of 3.0 GHz and a boost clock of 4.1 GHz. This particular CPU is expected to have clock speeds that are 400 MHz lower than the 5800X3D SKU.

[...] The current report does not offer details on the potential availability of the newly discussed CPUs. Right now, AMD has two new CPUs with 3D V-Cache. One is called the 5800X3D and is considered the best gaming CPU for the AM4 platform. The other is called the 5600X3D and can only be bought from MicroCenter in the United States. The prospect of AMD broadening the availability of 3D V-Cache options globally, rather than confining them to specific retailers, would indeed be noteworthy.

The introduction of the 5700X3D and 5500X3D models for the AM4 platform could make it better for gamers who don't want to spend too much money. The release of more cost-effective SKUs could offer gamers an attractive upgrade path without imposing a substantial financial burden, paving the way for a transition to newer AM5/LGA-1851 sockets in the future.

A noteworthy aspect of the X3D series is its compatibility with various entry-level motherboards. These CPUs do not support overclocking. Hence, their requirements are lower. Despite this limitation, these CPUs show superior performance in gaming tests compared to their non-X3D counterparts.

Source: @g01d3nm4ng0

Original Submission

taylorvich writes:

https://arstechnica.com/science/2023/11/capacitor-based-heat-pumps-see-big-boost-in-efficiency/

Various forms of heat pumps—refrigerators, air conditioners, heaters—are estimated to consume about 30 percent of the world's electricity. And that number is almost certain to rise, as heat pumps play a very large role in efforts to electrify heating to reduce the use of fossil fuels.

Most existing versions of these systems rely on the compression of a class of chemicals called hydrofluorocarbons, gasses that were chosen because they have a far smaller impact on the ozone layer than earlier refrigerants. Unfortunately, they are also extremely potent greenhouse gasses, with a short-term impact several thousand times that of carbon dioxide.

Alternate technologies have been tested, but all of them have at least one major drawback in comparison to gas compression. In a paper released in today's issue of Science, however, researchers describe progress on a form of heat pump that is built around a capacitor that changes temperature as it's charged and discharged. Because the energy spent while charging it can be used on discharge, the system has the potential to be highly efficient.

[...] For hydrofluorocarbons, the difference in heat content can be controlled by altering the pressure. Compressing a gas will heat it up while lowering the pressure cools it down. However, various other materials undergo similar heating and cooling in response to other external influences, including physical stress, magnetic fields, or electric fields. In many cases, these materials remain solid despite experiencing significant changes in temperature, which could potentially simplify the supporting equipment needed for heating and cooling.

In the new work, done by researchers mostly based in Luxembourg, the researchers focused on materials that change temperature in response to electric fields, generically known as electrocalorics. While a variety of configurations have been tested for these materials, researchers have settled on a layered capacitor structure, with the electric field of the material changing as more charge is stored within it. As charge is stored, an electrocaloric material will heat up. When the charge is drained, they'll draw in heat from the environment.

This has a significant advantage regarding the power needed for the device to operate since the current generated when draining the capacitor can just be used to power something. There's a little energy lost during the round-trip in and out of storage, but that can potentially be limited to less than one percent.

Journal Reference:
Junning Li et al., High cooling performance in a double-loop electrocaloric heat pump, Science, 16 Nov 2023 Vol 382, Issue 6672 pp. 801-805 DOI: 10.1126/science.adi5477

Original Submission #1  Original Submission #2

takyon writes:

OpenAI President Greg Brockman quits as shocked employees hold all-hands meeting

On Friday afternoon, not long after news of CEO Sam Altman's abrupt and surprising departure from OpenAI began spreading online, the company held an all-hands meeting at its headquarters in San Francisco, reports The Information. During the meeting, interim CEO Mira Murati attempted to reassure the shocked employees that the search for a new CEO is underway.

Hours later, OpenAI co-founder and president Greg Brockman posted a statement on X, saying that after he learned today's news he sent a message to the OpenAI team: "based on todays news, i quit." Brockman, a key technical figure involved in many of the company's successes, was relieved of his OpenAI board membership on Friday, but the company initially announced he would be staying on.

Earlier on Friday, OpenAI released a blog post titled "OpenAI announces leadership transition" where it announced that Atlman "was not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities." In a response post on X, Altman wrote, "I loved my time at openai," and hinted at future plans without revealing any details.

OpenAI co-founder and head of research Ilya Sutskever may have led the "coup". Other board members involved in the removal include Adam D'Angelo (Quora CEO), Tasha McCauley (RAND Corporation), and Helen Toner (Georgetown's Center for Security and Emerging Technology).

See also:

WTF is going on at OpenAI? We have theories
Why Was Sam Altman Fired As CEO of OpenAI?
Everything We Know About OpenAI's Shocking Ouster of Sam Altman

Freeman writes:

https://arstechnica.com/tech-policy/2023/11/apple-discriminated-against-us-citizens-in-hiring-doj-says/

Apple illegally discriminated against US citizens and other US residents in its hiring and recruitment practices for certain types of positions that went to foreign workers, the US Department of Justice said yesterday. Apple agreed to pay up to $25 million in back pay and civil penalties to settle the DOJ allegations.

Apple discriminated "against US citizens and certain non-US citizens whose permission to live in and work in the United States does not expire," the agency said. The $25 million payment was called the largest ever collected by the Justice Department under the anti-discrimination provision of the Immigration and Nationality Act (INA).

Apple is required to pay $6.75 million in civil penalties and create an $18.25 million fund to provide back pay to those harmed by its hiring practices. Apple did not admit guilt in the settlement. But the company acknowledged in a statement that it had "unintentionally not been following the DOJ standard," according to Reuters.

Original Submission

You may think you have fast fingers from typing or texting, but how fast can you assemble a V-8 NASCAR engine? Jayski (NASCAR news site) reports on the annual Hendrick Motorsports engine building contest, https://www.jayski.com/2023/11/14/danny-emerick-bill-sullivan-win-2023-randy-dorton-hendrick-engine-builder-showdown/

The team of Danny Emerick and Bill Sullivan edged out the team of Scott Vester and Phil Seaton by 0.91 [seconds] to win the 2023 Randy Dorton Hendrick Engine Builder Showdown. With the victory, Emerick joins Vester as a six-time winner of the annual competition.
The two teams were neck and neck coming to the finish with the Emerick-led team overcoming an early miscue to take the title. The Emerick-Sullivan team posted a time of 22:56.46, while the Vester-Seaton squad posted a time of 22:57.37.

[...] The Randy Dorton Hendrick Engine Builder Showdown sees 12 Hendrick Certified Master Technicians from all across the country at Hendrick Automotive Group, paired with 12 Hendrick Motorsports engine department team members. Each two-person team is matched up against another duo looking to post the fastest qualifying time. The two teams assemble 358-cubic-inch Chevrolet engines with 243 parts[1], similar to the fuel-injected engines that run in the NASCAR Cup Series. The builders assemble their engines on their own stage platform as they race against the clock to post the fastest time. Winners are determined by the quickest time with the fewest number of errors. The top two teams with the fastest times face off in the championship round. To date, the quickest time recorded in this competition was 21 minutes and 40 seconds in 2014.

243 parts in 22 minutes means over 10 parts per minute, or an average of 5 or 6 seconds, per part added to the engine. Each part put in the correct place and in some cases tightened up & torqued (with a power wrench or screwdriver).

[1] your AC contributor believes that some of the "243 parts" may in fact be assemblies that come to the engine builders pre-assembled and tested (thus a _fully_ disassembled engine might have more parts), but that is just a guess.

A quick google suggests that these engines use timing belts instead of chains--in the past the high parts-count of IC engines could be due to counting the individual links and rollers in the timing chain.

Original Submission

taylorvich writes:

Research suggests that your body may enjoy an exercise routine more than you do:

If you've ever started a new exercise plan or diet, you'll most likely have been told many times that consistency is the key to reaching goals. Now, scientists for the first time have found that sticking to a daily routine of when you work out could be just as important for bone and joint health.

A study out of The University of Manchester has looked at internal body clocks to see how exercising at the same time of day can potentially shield against bone and joint deterioration, protecting against injury and stave off age-related physical decline and associated conditions such as arthritis.

"Among the many health challenges, the age-related musculoskeletal decline – and its adverse consequences – is a major burden to individuals," said senior author Judith Hoyland, and spine/intervertebral disk expert from The University of Manchester. "We have identified a new clock mechanism underlying skeletal aging, which could have far-reaching impacts on understanding frailty and designing more efficient treatment timing of exercise and physiotherapy to maintain good skeletal health and mobility."

[...] "The daily 24-hour cycle that our bodies follow, such as our internal temperature dropping when we sleep and our blood pressure rising at certain times of day, is known as our circadian rhythm," said Lucy Donaldson, Director of Research & Health Intelligence at collaborating organization Versus Arthritis. "There are processes inside our body which keep this rhythm going, known as 'clocks,' which are all linked to our central body clock in the brain."

Studies have shown that if those other clocks are out of whack with our central timekeeper, it presents a higher risk of many health conditions including cardiovascular disease. Research suggests fat cells have their own biological clock, and the cardiovascular system's own ticker may explain the prevalence of morning heart attacks.

owl writes:

https://www.theregister.com/2023/11/15/righttorepair_ftc_petition/

iFixit and the Public Interest Research Group (PIRG) have teamed up to go straight to the US Federal Trade Commission with a rulemaking petition urging it to implement national right-to-repair rules.

The petition [PDF], sent to the FTC on Tuesday, calls on the Commission to implement new regulations under its Section 5 powers, which gives the agency enforcement authority over "​​unfair or deceptive acts or practices in or affecting commerce."

"The FTC has been a strong ally in protecting our right to fix everything we own," said Director of Sustainability for iFixit, Liz Chamberlain. "But for the FTC to be fully empowered to fix the things stopping us from fixing things, they need new rules. This petition for rulemaking aims to give the FTC the power they need to ensure that we can all fix all our things."

Under the FTC's Section 5 authority, PIRG noted, the FTC could require that consumable components and parts that commonly fail are made readily available through a product's lifespan, ensure consumers have the right to repair devices how and where they see fit, that key functions remain enabled after a manufacturer ends support for a product, and that independent repair shops aren't required to share customer data with manufacturers.

Original Submission

Freeman writes:

https://arstechnica.com/information-technology/2023/11/people-think-white-ai-generated-faces-are-more-real-than-actual-photos-study-says/

Eight images used in the study; four of them are synthetic. Can you tell which ones? (Answers at bottom of the article.)

A study published in the peer-reviewed journal Psychological Science on Monday found that AI-generated faces generated with three year-old technology, particularly those representing white individuals, were perceived as more real than actual face photographs, reports The Guardian. The finding did not extend to images of people of color, likely due to AI models being trained predominantly on images of white individuals—a common bias that is well-known in machine learning research.
[...]
Researchers used real and synthetic images sourced from an earlier study, with the synthetic ones generated by Nvidia's StyleGAN2 image generator, which can create realistic faces using image synthesis. It's worth noting that StyleGAN2 was released in 2020, and AI image synthesis has progressed rapidly since then—but newer AI models were not involved in the study.

Original Submission

upstart writes:

The space agency is waiting out a phenomenon known as a solar conjunction:

NASA is putting pause on sending commands to its Mars exploration instruments from November 11 through November 25 as it waits out the Mars solar conjunction. With the sun in the way, any commands sent to Mars could suffer interference capable of harming the robotic explorers.

NASA's Mars exploration robots will be on their own for the next two weeks while the space agency waits out a natural phenomenon that will prevent normal communications. Mars and Earth have reached positions in their orbits that put them on opposite sides of the sun, in an alignment known as solar conjunction. During this time, NASA says it's risky to try and send commands to its instruments on Mars because interference from the sun could have a detrimental effect.

That means the Perseverance and Curiosity rovers, the Ingenuity helicopter, the Mars Reconnaissance Orbiter, and the Odyssey and MAVEN orbiters will be left to their own devices for a little while. Their onboard instruments will continue to gather data for their respective missions, but won't send this information back to Earth until the blackout ends.

Original Submission

upstart writes:

Many transgressions come from "very large companies that have robust security teams":

Despite more than a decade of reminding, prodding, and downright nagging, a surprising number of developers still can't bring themselves to keep their code free of credentials that provide the keys to their kingdoms to anyone who takes the time to look for them.

The lapse stems from immature coding practices in which developers embed cryptographic keys, security tokens, passwords, and other forms of credentials directly into the source code they write. The credentials make it easy for the underlying program to access databases or cloud services necessary for it to work as intended. I published one such PSA in 2013 after discovering simple searches that turned up dozens of accounts that appeared to expose credentials securing computer-to-server SSH accounts. One of the credentials appeared to grant access to an account on Chromium.org, the repository that stores the source code for Google's open source browser.

In 2015, Uber learned the hard way just how damaging the practice can be. One or more developers for the ride service had embedded a unique security key into code and then shared that code on a public GitHub page. Hackers then copied the key and used it to access an internal Uber database and, from there, steal sensitive data belonging to 50,000 Uber drivers.

Uber lawyers argued at the time that "the contents of these internal database files are closely guarded by Uber," but that contention is undermined by means the company took in safeguarding the data, which was no better than stashing a house key under a door mat.

Freeman writes:

https://arstechnica.com/security/2023/11/teens-with-digital-bazookas-are-winning-the-ransomware-war-researcher-laments/

What do Boeing, an Australian shipping company, the world's largest bank, and one of the world's biggest law firms have in common? All four have suffered cybersecurity breaches, most likely at the hands of teenage hackers, after failing to patch a critical vulnerability that security experts have warned of for more than a month, according to a post published Monday.

[...] All four companies have confirmed succumbing to security incidents in recent days, and China's ICBC has reportedly paid an undisclosed ransom in exchange for encryption keys to data that has been unavailable ever since.

[...] After the CitrixBleed exploit grants initial remote access through software known as Virtual Desktop Infrastructure, LockBit escalates its access to other parts of the compromised network using tools such as Atera, which provides interactive PowerShell interfaces that don't trigger antivirus or endpoint detection alerts. This access remains even after CitrixBleed is patched unless administrators take special actions.

Original Submission

Arthur T Knackerbracket has processed the following story:

ChatGPT's recently-added Code Interpreter makes writing Python code with AI much more powerful, because it actually writes the code and then runs it for you in a sandboxed environment. Unfortunately, this sandboxed environment, which is also used to handle any spreadsheets you want ChatGPT to analyze and chart, is wide open to prompt injection attacks that exfiltrate your data. 

Using a ChatGPT Plus account, which is necessary to get the new features, I was able to reproduce the exploit, which was first reported on Twitter by security researcher Johann Rehberger. It involves pasting a third-party URL into the chat window and then watching as the bot interprets instructions on the web page the same way it would commands the user entered. 

[...] I tried this prompt injection exploit and some variations on it several times over  a few days. It worked a lot of the time, but not always. In some chat sessions, ChatGPT would refuse to load an external web page at all, but then would do so if I launched a new chat. 

In other chat sessions, it would give a message saying that it's not allowed to transmit data from files this way. And in yet other sessions, the injection would work, but rather than transmitting the data directly to http://myserver.com/data.php?mydata=[DATA], it would provide a hyperlink in its response and I would need to click that link for the data to transmit.

I was also able to use the exploit after I'd uploaded a .csv file with important data in it to use for data analysis. So this vulnerability applies not only to code you're testing but also to spreadsheets you might want ChatGPT to use for charting or summarization.

[...] The problem is that, no matter how far-fetched it might seem, this is a security hole that shouldn't be there. ChatGPT should not follow instructions that it finds on a web page, but it does and has for a long time. We reported on ChatGPT prompt injection (via YouTube videos) back in May after Rehberger himself responsibly disclosed the issue to OpenAI in April. The ability to upload files and run code in ChatGPT Plus is new (recently out of beta) but the ability to inject prompts from a URL, video or a PDF is not.

Original Submission

SpaceX's Starship gets all clear for Texas takeoff

DannyB writes:

FAA grants license to fly, though not fry local wildlife:

The US Federal Aviation Administration (FAA) has given SpaceX clearance to try launching the monster Starship / Super Heavy combo from the company's Boca Chica facility.

It has taken some time, at least in terms of the rapid iterative approach adopted by the company for its other vehicles. Still, SpaceX is set for another attempt seven months after April's effort.

Residing at the bottom of an FAA advisory are three possible dates for flight two of the SpaceX Starship Super Heavy. The primary date is November 17, with backup dates on November 18 and 19.

The FAA grounded SpaceX's Starship after the rocket demolished a chunk of its launchpad and scattered debris over the surrounding area. The launch was aborted a few minutes into flight, although there was a worrying delay between the red button being pushed and the tumbling rocket detonating.

See Also:
YouTube video: Musk Confirms License: Starship To Launch Friday!

Arthur T Knackerbracket has processed the following story:

A tool bag is orbiting Earth. No, this isn't an elaborate Elon Musk joke.

The bag entered orbit during a spacewalk conducted by NASA astronauts and International Space Station residents Jasmin Moghbeli and Loral O'Hara on November 1. During their almost seven-hour space stroll, during which they replaced bearings on a sun-tracking solar array and removed some communications equipment, NASA revealed the tool bag "was inadvertently lost," by one of the dynamic duo.

"Mission Control analyzed the bag's trajectory and determined that risk of recontacting the station is low and that the onboard crew and space station are safe with no action required," the space agency added.

Dr Meganne Christian, a member of the European Space Agency’s 2022 astronaut class, shared a snippet of helmet camera footage from Moghbeli's space suit showing the bag slipping away and the futile scramble to retrieve it.

But the container – a “crew lock bag” in official NASA parlance – didn’t just drift away out of human view.

Astrophysicist Jonathan McDowell of the Harvard Center for Astrophysics confirmed last week that the bag had been catalogued by the US Space Force as 58229/1998-067WC and was being tracked as a new orbital object.

The next day, Japanese ISS resident Satoshi Furukawa snapped a photo of the bag following day as the ISS passed over Japan.

According to McDowell, the bag isn't in a stable orbit, and is expected to re-enter Earth's atmosphere in a few months, when it will burn up completely and won't be a risk to anyone on the ground.

For those interested in observing the bag before it burns up in Earth's atmosphere, that just might be possible with the proper equipment and timing.

The bag's brightness is magnitude six , or just at the edge of the eye's unaided visibility limit under perfect conditions. Binoculars would make spotting it far easier. Sat-spotters seeking a sighting of the space spanners should look a few minutes ahead of the ISS’s expected path, which can be tracked online or with a recently launched mobile app.

Original Submission